Real-word outcomes for high-risk non-muscle-invasive bladder cancer: screened patients for the BRAVO trial.

OBJECTIVE To report real-world outcomes for high-risk non-muscle-invasive bladder cancer (HRNMIBC), including bacillus Calmette-Guérin (BCG) and radical cystectomy (RC), as randomised comparisons of these have not been possible. METHODS We detail consecutive participants screened for the BRAVO randomised controlled trial comparing RC with BCG (International Standard Randomised Controlled Trial Number [ISRCTN]12509361). Patients were prospectively registered and case-note review used for outcomes. The primary outcome was overall survival. Secondary outcomes included recurrence, progression, metastasis, and bladder cancer-specific survival. RESULTS AND LIMITATIONS A total of 193 patients were screened, including 106 (54.9%) who received BCG, 43 (22.3%) primary RC, 37 (19.2%) 'other' treatment and seven (3.6%) hyperthermic intravesical mitomycin C. All-cause death occurred in 55 (28.5%) patients at median (interquartile range [IQR]) of 29.0 (19.5-42.0) months. In multivariable analysis, overall mortality was more common in older patients (hazard ratio [HR] 2.63, 95% confidence interval [CI] 1.35-5.13; Cox P = 0.004 for age >70 years), those recruited from district hospitals (HR 0.53, 95% CI 0.3-0.95; P = 0.032) and those who did not undergo RC as their first treatment (HR 2.16, 95% CI 1.17-3.99; P = 0.014). In all, 17 (8.8%) patients died from bladder cancer (BC) at median (IQR) of 22.5 (19-36.25) months. In multivariable analysis, BC-specific mortality was more common in older patients (HR 4.87, 95% CI 1.1-21.6; P = 0.037) and those with Tis/T1 disease (HR 2.26, 95% CI 1.23-4.16; P = 0.008) but did not vary with initial treatment. CONCLUSIONS Patients with HRNMIBC are at high-risk of mortality. Those choosing RC as their initial treatment have lower risks of mortality than others, although this may reflect fitness and selection.